Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Portland State UniversityPDXScholar
Economics Faculty Publications and Presentations Economics
9-1-1999
Climate, Water Navigability, and Economic DevelopmentAndrew D. Mellinger
Jeffrey D. Sachs
John Luke GallupPortland State University
Let us know how access to this document benefits you.Follow this and additional works at: https://pdxscholar.library.pdx.edu/econ_fac
Part of the Growth and Development Commons
This Working Paper is brought to you for free and open access. It has been accepted for inclusion in Economics Faculty Publications and Presentationsby an authorized administrator of PDXScholar. For more information, please contact [email protected].
Citation DetailsMellinger, Andrew D., Jeffrey D. Sachs, and John L. Gallup. "Climate, water navigability, and economic development." (1999).
Climate, Water Navigability,and Economic Development
Andrew D. Mellinger, Jeffrey D. Sachs,and John L. Gallup
CID Working Paper No. 24September 1999
© Copyright 1999 Andrew D. Mellinger, Jeffrey D. Sachs, John L. Gallup,and the President and Fellows of Harvard College
Working PapersCenter for International Developmentat Harvard University
CID Working Paper no. 24
Climate, Water Navigability, and Economic Development
Andrew D. Mellinger, Jeffrey D. Sachs, and John L. Gallup*
Abstract
Geographic information systems (GIS) data was used on a global scale to examine therelationship between climate (ecozones), water navigability, and economic development in termsof GDP per capita. GDP per capita and the spatial density of economic activity measured as GDPper km2 are high in temperate ecozones and in regions proximate to the sea (within 100 km of theocean or a sea-navigable waterway). Temperate ecozones proximate to the sea account for 8percent of the world’s inhabited land area, 23 percent of the world’s population, and 53 percentof the world’s GDP. The GDP densities in temperate ecozones proximate to the sea are onaverage eighteen times higher than in non-proximate non-temperate areas.
JEL codes: O10 Economic Development – General, O 40 – Economic Growth and AggregateProductivity, R10 – General Spatial Economics
Keywords: geography, economic development, economic density, climate, GDP per capita
Andrew D. Mellinger is a Research Associate at the Center for International Developmentspecializing in the multi-disciplinary application of geographic information systems. His areas ofresearch include landscape ecology, analysis and planning, the modeling of urbanization andalternative futures, sustainable land-use planning, and comparing road networks in spatialecological change models. [email protected]
Jeffrey D. Sachs is the Director of the Center for International Development and the Galen L.Stone Professor of International Trade in the Department of Economics at Harvard. His currentresearch interests include emerging markets, global competitiveness, economic growth anddevelopment, transition to a market economy, international financial markets, internationalmacroeconomic policy coordination and macroeconomic policies in developing and developedcountries. [email protected]
John L. Gallup is a Research Fellow at the Center for International Development, an InstituteAssociate at the Harvard Institute for International Development, and a Lecturer in theDepartment of Economics at Harvard University. His recent research focuses on economicgeography, malaria, and the relationship between economic growth and poverty [email protected]
*This research is part of an ongoing project at Harvard’s Center for International Development investigating thelinkages between geography and economic development. The author’s wish to thank Karin Johnson, AndrewWarner, and Steve Radelet for their work on this paper and related efforts.
CID Working Paper no. 24
Climate, Water Navigability, and Economic Development
Andrew D. Mellinger, Jeffrey D. Sachs, and John L. Gallup
I. Introduction
One of the central issues of economics is the enormous disparity in economic
performance between rich and poor regions of the world. Modern economics got its start, in fact,
with Adam Smith’s Inquiry into the Wealth of Nations in 1776, in which Smith identified social
and geographical factors that could account for differential economic performance across regions
of the world. Smith is remembered today mainly for his theory that market institutions would
enable societies to develop a richer division of labor, and therefore higher living standards, than
societies subject to extensive government controls. He is less remembered for his equally astute
geographical observations. Smith asserted that the division of labor is limited by the extent of
the market, and that coastal regions, by virtue of their ability to engage in sea-based trade, enjoy a
wider scope of the market than interior regions. In Smith’s words:
As by means of water carriage a more extensive market is opened to every sort of industrythan what land carriage alone can afford it, so it is upon the sea-coast, and along thebanks of navigable rivers that industry of every kind begins to sub-divide and improveitself, and it is frequently not till a long time after that those improvements extendthemselves to the inland part of the country
Based on the importance of sea-based trade, Smith drew pessimistic conclusions regarding inland
Africa and large parts of Russia, Siberia and Central Asia:
All the inland parts of Africa, and all that part of Asia which lies any considerable waynorth of the Black and Caspian Seas, the ancient Scythia, the modern Tartary and Siberia,seem in all ages of the world to have been in the same barbarous and uncivilized state inwhich we find them at present. . . There are in Africa none of those great inlets . . . tocarry maritime trade into the interior parts of that great continent. . .
Smith was less aware of the effects of climate on economic performance. Climatic
conditions have pervasive effects on disease, agriculture, human physiology, and other factors
2
that may affect economic performance. Recent studies (Gallup, Sachs, and Mellinger, 1998;
Bloom and Sachs, 1998) have noted that tropical areas are consistently poorer than temperate
zone areas, because of the intrinsic effects of tropical ecology on human health and agricultural
productivity. Tropical infectious diseases, for example, impose very high burdens on human
health, that in turn may lead to shortfalls in economic performance much larger than their direct,
short-run effects on health. Another recent study (Gallup and Sachs 1999), found that after
controlling for material inputs such as capital, labor, and fertilizers, the productivity of tropical
food production still falls far short of the productivity of temperate zone food production.
In an overview of economic development and geography (Gallup, Sachs, and Mellinger,
1998) the latitudinal belt between the Tropics of Cancer (23.45 N) and Capricorn (23.45 S) was
used to separate the geographical tropics from the rest of the world. They showed that economies
in the geographic tropics have lower income levels and lower growth rates than the rest of the
world, and that the shortfall is evident after controlling for other standard economic variables that
affect economic performance. Their analysis is extended here by using ecological measures of
the tropics rather than the simple latitudinal definitions.
In the following sections, we examine the geographic distribution of per capita GDP,
GDP density (defined as GDP per km2 ), and population density. These variables are highly
influenced by climate and proximity to the sea. We find strong evidence that the ecological
tropics, the dry regions, and the sub-tropical regions are systematically poorer than temperate
ecozones. Moreover, the temperate ecozones proximate to the sea, though a small part of the
world’s inhabited landmass, account for a remarkably high proportion of the world’s annual
economic output. Section II describes the data set we developed for this research. Section III
describes the global distributions of GDP and population across the ecozones. Section IV
3
examines the distributions within continents. Section V offers some further discussion of the
results and presents conclusions and questions for investigation.
II. Data and Methods
A geographic information system (GIS) was assembled containing four variables: climate
zones, population, navigable rivers and Gross Domestic Product on a per capita basis for 152
countries with a population of 1 million or more in 1995. In 1995 these countries had a
combined population of 5.65 billion, 99.7 percent of the world’s population (Tobler, et al, 1995).
GIS is a computer-based relational database used for the storage, analysis, and display of
geographically referenced data. The inherent advantage to using a GIS is that data can be
analyzed spatially.
A digital map of climate was constructed using the classification devised by Wladmir
Köppen in 1918 and revised in 1953 by his students Geiger and Pohl to determine climate
boundaries that coincided with major vegetation types (Strahler and Strahler, 1992). The
usefulness of this approach to classification lies in its empirical delineation of climatic
boundaries based on either monthly or annual actual temperature and precipitation values. The
Köppen -Geiger-Pohl classification designates major climate groups, subgroups within the major
groups, and further divides the subgroups to capture seasonal differences in temperature and
precipitation.
The six major categories of climate zones are designated by a capital letter: A = tropical,
rainy; B = dry; C = mild, humid; D = snow, forest; E = polar; and H = highland. The subgroups
are classified by the addition of another letter. Adding the lower case f used with Af, Cf, and Df
indicates climates that are moist with adequate precipitation in all months with no dry season.
4
The w used with Aw and Cw denotes a dry winter season in the respective hemisphere. The m,
which is only used with the A climate, indicates a rainforest climate despite a short, dry season in
monsoon-like precipitation cycles. The s used with Cs indicates a dry season in the summer of
the respective hemisphere. The upper case W represents arid, desert climates and the S indicates
semiarid, steppe climates. The W and S are used only with the dry B climates. The H and E
zones do not have further subdivision. Figure 1 depicts the global extent of the subgroups. We
mainly are concerned with the eleven classifications of the second tier, Af, Am, Aw, BS, BW,
Cf, Cs, Cw, Df, DW, and H. The E climate is generally excluded because the 4 percent of the
world’s land area in this polar zone (e.g. the tundra of Northern Russia) has almost no human
population.
Figure 2 shows the global distribution of population density as of 1994 (Tobler, et al,
1995) measured as population per km2. Tobler et al (1995) obtained a resolution of 5 minutes by
5 minutes, approximately 7.5 km2 at the equator. Some of the underlying data, however, is less
refined, with population interpolated to the 5 minute by 5 minute grid. A detailed description of
the gridded population dataset is in the endnote reference.
Since we are concerned with modern economic development as driven by international
trade, rivers were mapped based on navigability for ocean-going vessels. Rivers categorized as
navigable in the ArcAtlas (ESRI, 1996) database are pared down according to three rules using
information taken from Rand McNally (1980), Brittanica Online (1998), and Encyclopedia
Encarta (1998): whether a river accommodates vessels with a minimum draft of approximately 3
meters (anything smaller is not considered ocean-going); the point at which a navigable river
becomes obstructed by falls, rapids, locks or dams; and whether the river is frozen during winter.
The coastline used is free from pack ice throughout the year (ESRI, 1996; Rand McNally, 1994).
5
Using this classification Figure 3 shows the land area in the world that is within 100 km of the
ocean coast or a sea-navigable river. As Adam Smith noted, Africa has no sea-navigable rivers
that extend from the oceans to the interior of the continent. High continental tableland precludes
navigability, even with investment. By contrast, North America has two navigable waterways
that link the continent’s vast interior with the ocean: the St. Lawrence Seaway and Great Lakes
system, and the Mississippi River system (and its major tributaries such as the Missouri and Ohio
Rivers)1.
Per capita Gross Domestic Product (GDP) was measured at standardized purchasing
power parity (PPP) at both the national and sub-national level. Figure 4 depicts global
distribution of GDP per capita in 1995. To capture intra-country variance in income distribution,
sub-national per capita GDP data was gathered for 19 of the 152 countries in our GIS, including
most of the large economies. These are Australia, Belgium, Brazil, Canada, Chile, China,
Colombia, France, Germany, Greece, India, Italy, Japan, Mexico, Netherlands, Spain, United
Kingdom, United States, and Uruguay. Since the sub-national data of these countries was
collected in local currency rather than in a comparable $US purchasing power parity basis, the
local-currency measures were adjusted to create internationally comparable sub-national
measures.2 The two major gradients of per capita GDP that result are climate and distance from
1 The Saint Lawrence Seaway is partly man-made, but the improvements upon its already largely navigable coursemade this river system fully passable to the Atlantic.2 The country-level $US PPP-adjusted GDP is used for each country for 1995 (GDPc below), available from theWorld Bank (1998) World Development Indicators, supplemented by CIA World Factbook (1996, 1997) estimates.For each sub-national region i, the $US PPP GDPi is calculated accordingly:
$US PPP GDPi = GDPc x (GDPi/GDPa)
GDPi is the GDP per capita of region i in local currency, and GDPa is the country-average per capita GDP in localcurrency. GDPa is calculated as Σ(GDPi x Popi)/Σ(Popi). Sub-national regional GDP was collected from nationalsources, mainly government statistical yearbooks. Thus, provincial populations and provincial income data are usedto calculate a country’s average GDP per capita, and then used to calculate the ratio of each province’s per capitaGDP to the national average. We then multiply that ratio by the $US PPP GDP, to calculate a GDP on a PPP basis
6
the coast. The tropical regions are almost all poor and coastal regions tend to have higher
incomes than interior regions. These effects are quantified in the next section.
III. Spatial distribution of population and economic density
Using the GIS population data, GDP per capita, and climate zone, the distribution of
economic activity and global population according to ecozones and proximity to the sea is
calculated. Tables 1-3 show the proportions of global land area, population, and total GDP
within each of the eleven climate zones. Each climate zone is separated into two sub-regions:
“near,” signifying within 100 km of the sea (i.e. ocean coast or ocean-navigable river); and “far,”
signifying beyond 100 km of the sea. Four sub-zones (Cf, Cs, Dw, and Df) are classified as
“temperate.” The Cw zone is mainly sub-tropical but is included here as part of the tropical zone
rather than the temperate zone (it encompasses the Gangetic valley of India, stretches in South
America south of the Amazon, in Southern Africa and a small belt in northeastern Australia).
The near temperate zone which is within 100km of the sea and in a temperate climate
plays a dominant role in the world economy, see Figure 5. Note that much of the United States’
coasts, Great Lakes, and Mississippi River are included, almost all of Western Europe, much of
East Asia including coastal China, South Korea, and Japan, coastal Australia, New Zealand,
South America’s Chile, coastal Argentina, and a small part of coastal Brazil, part of North
Africa’s coast, and the southern tip of South Africa. These regions contain almost all of the
world’s economic powerhouse economies, and as is demonstrated a significant proportion of
global production.
for each region. This calculation assumes that the ratio of the regional GDP per capita to national GDP per capita inlocal currency equals the ratio of the regional GDP to national GDP on a PPP basis.
7
Table 1 shows the proportion of global land area in each ecozone. Note that 17.4 percent
of the world’s land area is within 100 km of the sea. 39.2 percent of the world’s land area lies
within the four sub-zones designated “temperate.” By calculating the intersection of the two
areas, the near temperate zone constitutes 8.3 percent of the world’s land. Nearly one third of the
world’s land area (29.6 percent) lies in the dry climate zones (desert and steppes), along with a
smaller proportion of the world’s population (18.0 percent). The dry zones tend to be among the
least densely populated places on the planet. 19.9 percent of the world’s land area lies within the
tropics (Af, Am, Aw and Cw), and 7.3 percent within the populated highlands.
Table 2 repeats this exercise for world population. 34.9 percent live in the temperate
zone, and 49.9 percent live within 100 km of the sea. The near temperate region includes 22.8
percent of the world’s population, located on 8.4 percent of the world’s landmass, making it a
densely populated part of the world. 24.3 percent live in the tropical A zones, while another 16.0
percent of the world, mainly in India and China, live in the sub-tropical Cw zone. 6.2 percent of
the world’s population lives in desert (BW) regions, which comprises 17.3 percent of the land.
Table 3 shows the spatial distribution of the world’s GDP by climate zone. For each sub-
region, per capita GDP (measured at PPP) is multiplied by the population, giving the total GDP
for the sub-region. The world’s GDP is the sum of these sub-regional GDPs. A striking 67.6
percent of the world’s GDP is produced within 100 km of the sea, though that area comprises
only 17.4 percent of the world’s landmass. 67.2 percent of the world’s GDP is produced in the
temperate climates, though these account for only 39.2 percent of the world’s landmass. The
near temperate region, with 8.3 percent of the world’s landmass and 22.8 percent of the world’s
population, produces a remarkable 52.9 percent of the world’s GDP.
8
By dividing the cells of Table 2 by those of Table 1, shown in Table 4, the population
density of the climate zones is shown relative to the global average population density, which is
42.5 inhabitants per km2 (Tobler et al, 1995). Thus, the Cf-near zone has a relative population
density of 3.15, or a population density 3.15 times the world average, which is 133.9 people per
km2 (= 3.15 x 42.5). The overall near temperate region is densely populated, with a relative
density of 2.72. The near zones are more densely populated than the far zones in every climate
zone.
Dividing the cells of Table 3 by those of Table 2, shown in Table 5, the GDP per capita
relative to the world average (or $5,500 at PPP) is presented. The GDP per capita shows two
systematic gradients: the near regions are higher than the far regions for every ecozone averaging
1.4 times the world average in the near regions and .7 times the world average in the far regions.
And the temperate ecozones’ GDP per capita is higher than the non-temperate (except in the DW
zone which mostly encompasses Siberia) averaging 1.9 times the world average in the temperate
zones and .5 times the world average in the non-temperate zones. The highest total income
ecozone is the Cf (mild, temperate) climate, followed by the Cs (Mediterranean) climate and the
Df (snow) climate (the order of these three changes when you compare near and far GDP-per
capita). Per capita GDP is especially high in the regions that are both temperate and proximate to
the sea, with 2.32 times the world average or $12,760 in PPP currency units (2.32 x $5,500).
The near temperate zone has 6 times the per capita GDP of the far tropical zone.
GDP density, measured as total GDP per km2, is a useful measure for understanding
where overall production of goods and services takes place. Since GDP density is equal to per
capita GDP multiplied by population density, and since both GDP per capita and population
density are especially high in the near temperate zones, the GDP density is extremely high in
9
those regions. Table 6 shows GDP density calculated by dividing the cells of Table 3 by those of
Table 1. The GDP density in the Cf near zone is a remarkable 7.63 times the world average GDP
density equaling $230,000 per km2. The highest income densities are found in the near zones,
the temperate climates, and also the sub-tropical Cw climate. The latter ecozone is characterized
by a relatively low GDP per capita, but an extremely high population density. The near ecozones
have on average 10 times the GDP densities of the far zones. The near temperate ecozones
produce income densities that average 18 times that of the far non-temperate ecozones.
Thus far averages have been examined for population density, GDP per capita, and GDP
per km2 by region. Sharp differences were found according to coastal proximity and ecozone.
But how significant are these differences? While formal statistical tests of differences of means
are not readily viable (there being little justification in assuming that the variables are drawn
from any particular underlying distribution), the overall distributions of the variables were
examined to assess how different these distributions are across regions. Table 7 shows the
distribution of GDP per capita by ecozone in both the near and far regions. The top half of the
table refers to regions proximate to the sea, and the bottom half refers to regions far from the sea.
For each ecozone, the percentage of the population living at each level of GDP per capita3 is
calculated.
Differences in per capita income across ecozones are reflected in sharp differences in the
overall distribution, not just in the means. Among the near regions, for example, no more than 1
percent of the populations in tropical regions (Af, Am, Aw, or Cw) are in the high-income
3 Note that we have GDP per capita on a national basis for most countries (133 with population of one million ormore), and a sub-national (provincial or state) basis for 19 countries, giving us 455 administrative units in total forwhich we have an estimate of average GDP per capita. To make the calculations for Tables 3, 5, and 7, we assumethat the entire population within each administrative unit has the average GDP per capita of that unit. Thus, weignore income inequality within administrative units.
10
category ($16,000 or more), while 47 percent of the populations of the temperate regions are in
the high-income category. The tropical regions are nearly uniformly poor. While temperate
regions have a wide income range with a small proportion (7 percent) of the temperate-zone
populations at income levels below $2,000, compared with 42 percent of the tropical zone
population.
The same calculations are made for distribution of population density by ecozone, shown
in Table 8, again dividing the near and far regions for separate analysis. One systematic gradient
is that the near ecozones are uniformly more densely populated than the far ecozones. There is
much less homogeneity of population density within the tropical and temperate ecozones. The
ecozones are less defining of population density than they are of GDP per capita. The tropical
zones display regions of both high population densities, 320+ per km2, and low population
densities, 0-20 per km2. The same is true of temperate ecozones.
IV. Continental Patterns
As a final exercise in this section, the allocation of populations of major continental
regions into ecozones and coastal proximity is shown in Table 9. The near temperate regions
were noted earlier for having much higher levels of income per capita than all other categories.
The continental regions differ markedly in the shares of their populations and land areas that fall
within these advantaged or disadvantaged zones. We divide the world into nine broad regions:
North America, Latin America, Western Europe, Middle East and North Africa, Sub-Saharan
Africa, East Asia, South Asia, Oceania, and Eastern Europe and Former Soviet Union.
Four continental regions have large populations living in non-temperate climates; South
Asia has no temperate climate, 96 percent of Sub-Saharan Africans, 88 percent of Latin
11
Americans and 70 percent of Middle Eastern and North Africans live in non-temperate climates.
The global average shows 65 percent of population residing in non-temperate climates. The
global average for population living in non-temperate far regions is 38 percent. The same four
continental regions again have higher shares, 78 percent of Sub-Saharan Africans, 59 percent of
South Asians; 48 percent in Latin Americans, and 45 percent of the Middle Eastern and North
Africans live in non-temperate far regions. In Latin America, one-fifth of the population lives in
a highland (H) climate far from the coast.
The people living on the other five continental regions are distributed mostly among the
temperate climates. East Asia is included in this grouping because its non-temperate far
population of 25 percent is lower than the global average of 38 percent. East Asia also has an
advantage with the majority of its population living in a region near to the sea. Eastern Europe
and the Former Soviet Union have 74 percent of the population in temperate climates, but only
33 percent in a near temperate zone. For Russia alone, the population share in the near temperate
region is only 6 percent.
Western Europe, North America, and Oceania are all especially favored. Western Europe
has a striking 96 percent of the population in a temperate ecozone, with 87 percent in near
temperate ecozones. North America has 88 percent of the population in temperate ecozones and
63 percent of the population in near temperate ecozones. In Oceania, 74 percent of the population
lives in the temperate ecozones, with 63 percent of the population in near temperate ecozones.
Figures 1 and 5 show that the continents with the highest concentrations of near temperate
populations, Western Europe, North America and Oceania, are also the richest. Conversely, the
two continents with the highest population in far tropical ecozones, Sub-Saharan Africa and
South Asia, are also the poorest.
12
V. Discussion and Future Research
Climate and coastal proximity are two key geographical gradients of economic
development. Temperate ecozones and regions within 100 km of sea navigable waterways are
home to more than 50 percent of the world’s economic output, but encompass only 8 percent of
the world’s inhabited landmass. When population is factored in the inequality is even greater.
The near ecozones contain on average ten times the GDP densities of the far ecozones.
Comparing the economic density of the near temperate ecozones with the far non-temperate eco-
zones, the GDP densities are on average 18 times greater. It is the task of the science of economic
development to give an interpretation of these patterns.
Three hypotheses seem appropriate for further exploration. The simplest hypothesis is
that the intrinsic characteristics of the tropics and interior regions are indeed highly detrimental to
long-term economic development. Tropical climates are burdened by much higher levels of
infectious disease than temperate climates, and are generally less productive in food production.
Interior regions suffer from much higher transport costs than coastal regions. The combination
of being both interior and non-temperate is therefore doubly detrimental. Sub-Saharan Africa
has no less than 78 percent of its population living in far and non-temperate regions.
A second hypothesis is that tropical climates are detrimental, but only modestly so. If the
world is subject to increasing-returns-to-scale production technologies, however, then small
initial disadvantages can cumulate into larger and larger differences over time. An example
might be the following. Suppose that tropical climates were only, say, 25 percent disadvantaged
200 years ago, but that innovative activity is ecozone-specific (for example in health and
agricultural technology) and is determined by the size of the market. A small initial advantage in
13
the temperate zone could then multiply as a result of much larger induced innovative activity in
the temperate zones as a result of the initial modest advantage. In this case, the main policy
implication would be the importance of re-directing scientific and technological efforts towards
tropical ecozone problems. This position is advocated in Sachs (1999).
A third kind of hypothesis would hold that the technological disadvantages of the tropics,
or of interior regions, is a thing of the past; that the disadvantages were once important, but no
longer are. In this case, the major differences in income levels across regions would tend to
diminish over time, except to the extent that increasing-returns-to-scale processes (such as
innovative activity) continue to magnify the former disadvantages into permanent differences.
We have been beginning some of this work, by examining the role of climate in the
process of technological innovation and technological diffusion. It seems, for example, to be the
case that innovative efforts in public health are still overwhelmingly directed at “temperate-zone
diseases,” and that the resulting technological innovations do not always cross the ecological
divide. There is remarkably little research on malaria vaccines, for example, even though the
technological barriers could likely be overcome in just a few years (Hamoudi, Kremer, and
Sachs, 1999). We have also found that the growth in total factor productivity in agriculture was
considerably higher in the temperate zones than the tropical zones during the period 1961 – 94
(See Gallup and Sachs, 1999).
With regard to coastal proximity, it might be supposed that the vast cost reductions in air
and land travel, and in telecommunications and data transmission, this century would have
greatly reduced the advantages of a coastal location. Such does not seem to be the case,
however. In the United States, for example, the proportion of the population living near the sea
coast has been rising steadily this century (Rappaport and Sachs, 1999). Also, coastal proximity
14
has given developing countries a clear advantage in the past thirty years in establishing
competitive manufacturing export sectors, which in turn have been important contributors to
overall economic growth (Radelet and Sachs, 1999).
15
References
Bloom, David and Jeffrey Sachs. 1998. "Geography, Demography, and Economic Growth inAfrica." Center for International Development, Cambridge, Mass. Mimeo.
Britannica Online. 1998. http://www.eb.com.
CIA. 1996. The World Factbook. Washington, D.C.: Central Intelligence Agency.
CIA. 1996. The World Factbook. http://www.odci.gov/cia/publications/
Encyclopaedia Encarta. http://encarta.msn.com.
ESRI (Environmental Systems Research Institute). 1992. ArcWorld’s User Guide and DataReference. Redlands, Calif.
ESRI. 1996a. ArcAtlas: Our Earth. Redlands, Calif.
ESRI. 1996b. ArcWorld Supplement Data Reference and User’s Guide. Redlands, Calif.
Gallup, John L., Jeffrey D. Sachs, and Andrew D. Mellinger. 1998. “Geography and EconomicDevelopment.” Annual World Bank Conference on Development Economics: 1998.Edited by Boris Pleskovic and Joseph E. Stiglitz. The World Bank, Washington, D.C.
Gallup, John L. and Jeffrey D. Sachs. 1999. “Agricultural Productivity and Geography.” Centerfor International Development, Cambridge, Mass. Mimeo.
Hamoudi, Amar, Michael Kremer, and Jeffrey D. Sachs. 1999. “The Case for a Vaccine PurchaseFund.” Mimeo. http://cid.harvard.edu/cidmalaria/
Köppen, Wladmir. 1918. “Klassifikation der Klimate nach Temperatur, Niederschlag andJahreslauf,” Petermanns Geographische Mitteilungen. Volume 64, pp. 193-203, 243-48.
Rappaport, Jordan and Jeffrey Sachs. 1999. “The United States as a Coastal Nation.” Mimeo.
Radelet, Steven and Jeffrey Sachs. 1999. "Shipping Costs, Manufactured Exports, and EconomicGrowth." Presented at the American Economics Association annual meeting, January1998.
Rand McNally and Co. 1980. Encyclopaedia of World Rivers, New York: Rand McNally andCo.
Sachs, Jeffrey D. 1999. “Helping the World’s Poorest.” The Economist. 352:8132:17-20.
Smith, Adam. 1976. An Inquiry into the Nature and Causes of the Wealth of Nations. 2 Vols.Edited by Edwin Canaan. Chicago, Ill.: Univ. of Chicago Press.
16
Strahler, Alan H. and Arthur N. Strahler. 1992. Modern Physical Geography. Fourth Edition.John Wiley and Sons, Inc. New York, New York. Map originally published in Geiger, R.and W. Pohl, Revision of the Köppen -Geiger Klimakarte der Erde (Darmstadt: JustusPerthes, 1953), 4 sheets, 1:16 Million: R. Geiger, “Eine neue Wandkarte derKlimagebiete der Erde,” Erdkunde, Vol. 8 (1954), pp. 58-61.
Tobler, Waldo, Uwe Deichmann, John Gottsegen, Kelley Maloy, “The Global DemographyProject” Technical Report 95-6, National Center for Geographic Information andAnalysis, University of California, Santa Barbara, CA, 1995. The data are also freelyavailable on the World Wide Web at CIESIN, “Gridded Population of the World,”http://www.ciesin.org.html
World Bank. 1998. World Development Indicators 1998 CD-ROM. Washington, D.C.:International Bank for Reconstruction and Development.
17
TABLE 1. LAND AREA BY CLIMATE ZONE
Near Far TotalAf 1.7% 2.3% 4.0%Am 0.7% 0.1% 0.8%Aw 2.5% 8.3% 10.8%Cw 0.6% 3.7% 4.3%BS 1.1% 11.2% 12.3%BW 1.9% 15.4% 17.3%H 0.4% 6.9% 7.3%E 0.1% 3.9% 4.0%Cf 4.8% 2.9% 7.7%Cs 1.3% 0.8% 2.2%Df 2.0% 21.0% 23.0%
DW 0.2% 6.2% 6.4%
Tropical1 6.1% 38.6% 44.7%Non-temperate2 9.0% 51.8% 60.8%
Temperate3 8.4% 30.9% 39.2%Total 17.4% 82.6%
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
Climate zoneLand area
18
TABLE 2. POPULATION BY CLIMATE ZONE
Near Far TotalAf 3.8% 0.6% 4.4%
Am 2.3% 0.1% 2.4%Aw 9.3% 8.2% 17.5%Cw 6.4% 9.6% 16.0%BS 2.3% 9.4% 11.8%BW 2.1% 4.1% 6.2%H 0.9% 5.9% 6.8%E 0.0% 0.0% 0.0%Cf 15.0% 4.5% 19.5%Cs 3.6% 0.7% 4.3%Df 2.7% 3.1% 5.8%
DW 1.5% 3.8% 5.3%
Tropical1 21.8% 18.5% 40.3%
Non-temperate2 27.1% 38.0% 65.1%
Temperate3 22.8% 12.1% 34.9%Total 49.9% 50.1%
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
Climate zonePopulation
19
TABLE 3. GDP BY CLIMATE ZONE
Near Far TotalAf 2.5% 0.3% 2.8%Am 1.0% 0.0% 1.0%Aw 3.6% 3.0% 6.6%Cw 3.4% 3.6% 7.0%BS 1.9% 4.6% 6.5%BW 1.4% 2.2% 3.6%H 0.9% 4.4% 5.3%Cf 36.3% 7.4% 43.7%Cs 7.9% 1.1% 9.1%Df 7.3% 3.7% 11.0%
DW 1.4% 2.0% 3.4%
Tropical1 10.5% 6.9% 17.4%
Non-temperate2 14.7% 18.1% 32.8%
Temperate3 52.9% 14.3% 67.2%Total 67.6% 32.4%
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
Climate zoneGDP
20
Near Far TotalAf 2.17 0.28 1.10
Am 3.56 0.62 3.16Aw 3.67 0.99 1.62Cw 11.03 2.57 3.70BS 2.19 0.84 0.96BW 1.12 0.27 0.36H 2.23 0.85 0.93Cf 3.15 1.57 2.55Cs 2.66 0.90 1.99Df 1.36 0.15 0.25
DW 6.05 0.61 0.82
Tropical1 3.59 0.48 0.90
Non-temperate2 3.01 0.73 1.07
Temperate3 2.72 0.39 0.89Total 2.87 0.61
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
TABLE 4. POPULATION DENSITY BY CLIMATE ZONE
Climate zonePopulation
21
TABLE 5: GDP PER CAPITA (GDP/pop)
Near Far TotalAf 0.66 0.54 0.64Am 0.41 0.30 0.41Aw 0.39 0.36 0.38Cw 0.54 0.37 0.44BS 0.80 0.49 0.55BW 0.65 0.54 0.58H 1.01 0.75 0.78Cf 2.42 1.63 2.24Cs 2.22 1.51 2.10Df 2.67 1.22 1.90
DW 0.92 0.53 0.64
Tropical1 0.48 0.37 0.43
Non-temperate2 0.54 0.48 0.50
Temperate3 2.32 1.18 1.94Total 1.35 0.65
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
Climate zoneGDP per capita
22
TABLE 6: GDP DENSITY (GDP/km2)
Near Far TotalAf 1.42 0.15 0.70
Am 1.46 0.19 1.28Aw 1.43 0.36 0.61Cw 5.91 0.95 1.61BS 1.75 0.41 0.53BW 0.73 0.15 0.21H 2.26 0.64 0.73Cf 7.63 2.56 5.71Cs 5.91 1.37 4.18Df 3.63 0.18 0.48
DW 5.57 0.33 0.53
Tropical1 1.90 0.48 0.87
Non-temperate2 1.63 0.35 0.53
Temperate3 6.32 0.46 1.72Total 3.89 0.39
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
Climate zoneGDP Density
23
TABLE 7. GDP PER CAPITA DISTRIBUTION% of population within each zone
Near $0-1,000 $1,000-2,000 $2,000-4,000 $4,000-8,000 $8,000-16,000 $16,000+Af 16% 6% 59% 8% 10% 1%
Am 4% 40% 52% 3% 0% 0%Aw 29% 37% 21% 12% 1% 1%Cw 25% 15% 34% 20% 6% 0%BS 2% 21% 55% 8% 9% 5%BW 11% 11% 50% 15% 11% 2%H 5% 3% 52% 22% 8% 10%Cf 0% 5% 15% 21% 6% 53%Cs 1% 2% 11% 30% 24% 33%Df 0% 1% 13% 30% 0% 56%
DW 22% 0% 15% 39% 24% 0%
Tropical1 23% 26% 35% 13% 4% 1%
Non-temperate2 19% 23% 38% 13% 5% 1%
Temperate3 2% 3% 14% 24% 9% 47%
Far $0-1,000 $1,000-2,000 $2,000-4,000 $4,000-8,000 $8,000-16,000 $16,000+Af 39% 9% 18% 20% 7% 7%
Am 3% 67% 30% 0% 0% 0%Aw 35% 37% 9% 12% 3% 3%Cw 27% 29% 29% 9% 3% 3%BS 21% 26% 37% 12% 2% 2%BW 11% 47% 19% 8% 7% 7%H 24% 26% 12% 29% 5% 5%Cf 1% 10% 48% 21% 10% 10%Cs 2% 3% 32% 30% 16% 16%Df 0% 0% 8% 91% 0% 0%
DW 1% 15% 59% 14% 5% 5%Tropical 31% 32% 20% 11% 3% 3%
Non-temperate 25% 31% 23% 13% 4% 4%Temperate 1% 9% 41% 36% 6% 6%
TOTALTropical 26% 29% 28% 12% 4% 2%
Non-temperate 23% 28% 29% 13% 4% 3%Temperate 2% 5% 23% 28% 8% 34%
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
24
TABLE 8. POPULATION DENSITY DISTRIBUTION (pop per km2)% of regional population with:
Near 0-20 20-40 40-80 80-160 160-320 320+Af 4% 7% 11% 17% 13% 48%
Am 3% 1% 5% 14% 26% 51%Aw 2% 3% 7% 7% 15% 67%Cw 0% 0% 1% 3% 8% 88%BS 3% 2% 14% 15% 18% 49%BW 9% 8% 8% 7% 13% 56%H 4% 7% 13% 26% 16% 34%Cf 2% 2% 7% 16% 18% 54%Cs 1% 3% 12% 25% 24% 35%Df 4% 5% 25% 25% 13% 27%
DW 2% 0% 1% 11% 20% 66%
Tropical1 2% 3% 7% 13% 17% 57%Non-temperate2 3% 3% 7% 9% 14% 65%
Temperate3 2% 3% 10% 18% 18% 49%
Far 0-20 20-40 40-80 80-160 160-320 320+Af 42% 16% 20% 13% 7% 2%
Am 28% 11% 7% 2% 11% 42%Aw 14% 9% 14% 22% 27% 14%Cw 4% 3% 6% 16% 25% 46%BS 11% 7% 12% 11% 19% 39%BW 24% 12% 7% 16% 16% 25%H 10% 12% 19% 22% 13% 24%Cf 5% 8% 14% 20% 21% 32%Cs 6% 20% 33% 28% 11% 3%Df 27% 20% 29% 11% 3% 10%
DW 6% 3% 8% 18% 25% 39%Tropical 12% 9% 14% 21% 24% 20%
Non-temperate 12% 8% 12% 17% 21% 31%Temperate 12% 11% 18% 17% 16% 26%
TOTALTropical 5% 5% 9% 15% 19% 46%
Non-temperate 8% 6% 10% 14% 18% 45%Temperate 6% 6% 13% 18% 18% 41%
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
25
TABLE 9. POPULATION BY CONTINENT
Continent Near Far Near Far Near Far Near FarSub-Saharan Africa 15% 47% 18% 78% 3% 1% 21% 79%
Eastern Europe and Former Soviet Union 0% 0% 6% 20% 33% 41% 39% 61%South Asia 38% 32% 41% 59% 0% 0% 41% 59%
Latin America 31% 25% 40% 48% 5% 7% 45% 55%Middle East & North Africa 0% 0% 25% 45% 23% 7% 48% 52%
Global average 22% 18% 27% 38% 23% 12% 50% 50%East Asia 28% 14% 32% 25% 26% 17% 58% 42%
North America 1% 0% 3% 8% 63% 26% 67% 33%Oceania 15% 2% 17% 8% 66% 8% 83% 17%
Western Europe 0% 0% 2% 2% 87% 9% 89% 11%
1) Tropical = Af, Am, Aw, and Cw
2) Non-temperate = Tropical + BS, BW, H and E
3) Temperate = Cf,Cs, Df, & DW
Tropical1 Non-temperate2 Temperate3 Total
26
Climate zonesAf (tropical rainforest climate)Am (monsoon variety of Af)Aw (tropical savannah climate)BS (steppe climate)BW (desert climate)Cf (mild humid climate with no dry season)Cs (mild humid climate with a dry summer)Cw (mild humid climate with a dry winter)DW (snowy-forest climate with a dry winter)Df (snowy-forest climate with a moist winter)E (polar ice climate)H (highland climate)
Tropic of Cancer
Tropic of Capricorn
Figure 1. Koppen-Geiger climate zones
27
Figure 2. Population distribution
Persons per square kilometer>11 - 23 - 56 - 1011 - 2021 - 5051 - 100101 - 34530
28
Figure 3. Land within 100 km of an ice-free coast or sea-navigable river
100 km zoneSea-navigable rivers
29
GDP PPP 1995 in $US$465 - 1,000$1,001 - 2,000$2,001 - 3000$3,001 - 5,000$5,001 - 8,000$8,001 - 12,000$12,001 - 20,000$20,001 - 44,000No Data
Tropic of Capricorn
Tropic of Cancer
Figure 4. GDP-PPP 1995 in $US
30
Tropic of Capricorn
Tropic of Cancer
Figure 5. GDP-PPP 1995 in $US in temperate climate zones 0-100km from the coast and sea-navigable rivers
GDP per square kilometer$ 0 - 499$ 500 - 1,099$ 1,100 - 2,999$ 3,000 - 8,099$ 8,100 - 21,199$ 22,000 - 59,999$ 60,000 - 162,999$ 163,000 - 441,999$ 442,000 - 546,000,000